RE: SKOS properties

> Quentin Reul wrote:
> > Hi all,
> > 
> > I think it would make sense to have a skos:antonym and it's quite
> > appropriate to think of two antonyms as being a special type of 
> > relations between terms. WordNet, which is probably one of the most 
> > common general thesaurus, uses antonyms. For example, the 
> antonym of 
> > "serviceman" is "civilian" but both are related as they 
> both describe a 
> > person.
Antonyms are not at all straightforward, in my view. Some other users
might think the antonym of "serviceman" was "armed robot", on the
grounds that a machine is the opposite of a human; others might even see
the opposite as being "servicewoman"! 
> > 
> > As SKOS is still in its development I don't why it wouldn't be 
> > possible
> > to add this feature to get a better linguistic 
> representation of thesaurus.
> 
> The problem is that with every new SKOS construct 
> we increase the complexity of SKOS and therefore 
> raise the barrier to its use. 

I agree this is an important consideration.


> So the question is 
> whether skos:antonym is sufficiently general to be 
> of use to tip the balance. I tend to agree with 
> Bernard that for the moment I would opt for the 
> local-extension approach. I expect the 
> documentation to contain lots of examples of how 
> to extend SKOS for typical cases such as antonym. 

I prefer this approach too. Thus it leaves the flexibility for thesaurus
managers to treat some antonyms as pseudo-synonyms and others as
associatively related preferred terms, according to the merits of each
case.

> BTW the fact that antonym is part of WordNet is 
> not really sufficient for me. There are some 17 
> relations in WordNet and SKOS will not support all 
> of them, I assume.
> 
> Guus
> [co-chair SWD]

Regards
Stella
*****************************************************
Stella Dextre Clarke
Information Consultant
Luke House, West Hendred, Wantage, Oxon, OX12 8RR, UK
Tel: 01235-833-298
Fax: 01235-863-298
SDClarke@LukeHouse.demon.co.uk
*****************************************************

Received on Monday, 30 April 2007 08:11:29 UTC